Short answer: Pick by what you do all day. If you write code, DeepLearning.AI's short developer course is the quickest structured start. If you don't, Google's Prompting Essentials on Coursera or Claude Academy's free courses fit better. And if a course doesn't make you write prompts against your own work, free lessons plus practice will teach you the same thing faster.
Disclosure: I have not taken any of these courses, and none of the links on this page are affiliate links. Everything below comes from each provider's own page, read on 23 Oct 2026. I compare fit, not teaching quality, because I can't judge quality without doing the work. MasterPrompting.net has its own free lessons, which I mention at the end, so weigh my view accordingly.
Searching "prompt engineering course" returns a wall of paid certificates, most with the same outline. Four options come from organisations people recognise and are worth checking before you pay anyone.
What does each course say about itself?
| Course | Format and length | Level | Price wording on its page |
|---|---|---|---|
| ChatGPT Prompt Engineering for Developers (DeepLearning.AI) | 1h40m, 9 video lessons, 7 code examples, 1 graded assignment | Beginner, basic Python needed | "Free for a limited time" during a beta; the graded assignment requires PRO |
| Google Prompting Essentials (Coursera) | "4 hours to complete", also "under 10 hours", four courses | Beginner, no prerequisites | "Included with Coursera Plus", $49/month after a 7-day free trial (US/Canada), financial aid mentioned |
| Claude Academy (AI Fluency and others) | Three courses listed: 4 hr, 3.5 hr and 45 min, each with lessons and a quiz | Not stated | Described as "Free courses, tutorials, and use cases" |
| Learn Prompting | Course list incl. Introduction to Prompt Engineering and Advanced Prompt Engineering; durations not on the list page | Not stated per course | Free plan (free courses, certificates of completion), Plus at $21/month billed annually |
Notes that matter:
- DeepLearning.AI is taught by Isa Fulford (OpenAI) and Andrew Ng, covers summarizing, inferring, transforming and expanding text, and builds a custom chatbot with the OpenAI API. The page shows "Updated Oct 12, 2026" and doesn't name the model it uses. "Free for a limited time" means the price can change, so don't plan around it.
- Google's course contradicts itself on price: the enrol button says "Enroll for free" while an FAQ says you can't take it for free. Read the checkout before you assume. It teaches a 5-step prompting framework through everyday tasks such as emails, data analysis and presentations, and its fourth course covers meta-prompting. It carries no university credit.
- Claude Academy is Anthropic's education site. Its three courses are about working with AI generally (fluency, capabilities and limitations, human-agent teams), not about tricks for one model, though it's Anthropic's own material. The page shows no certificate.
- Learn Prompting's list page gives no per-course price, duration or level, so I can't tell you which courses sit behind Plus.
Which one for which person?
| You are | Start with | Why |
|---|---|---|
| A developer who wants to call an LLM API | DeepLearning.AI developer course | Short, code-based, covers API use |
| A non-technical professional | Google Prompting Essentials | Everyday tasks, no prerequisites |
| Curious how to think about working with AI at all | Claude Academy courses | Free, concept-first |
| Someone who wants a long course library and a free tier | Learn Prompting | Free plan plus optional paid library |
| On a tight budget | Free options above, then practice | Cost is not the bottleneck, practice is |
One practical wrinkle for developers in India: the DeepLearning.AI course has you write prompts against an LLM API. I can't tell from its page whether the platform supplies the model access or you bring your own key. If you need your own, many Indian cards fail on provider billing pages. AICredits lets you top up in rupees from ₹50 and use an OpenAI-compatible endpoint (affiliate link), and a zero-forex card is the cheaper per-token option, compared in the credit card guide.
How do you judge any prompt engineering course?
Before you pay, check these:
- Does it make you write prompts for your own tasks, or only watch? Watching doesn't build the skill.
- Is the content dated? Look for an update date and for advice that no longer applies (long lists of "magic phrases", for example).
- Does it teach evaluation? Comparing outputs and testing changes matters more than any single technique. See evaluation frameworks.
- What does the certificate prove? I found no verified evidence that these certificates change hiring outcomes. A portfolio of working projects is easier to judge.
- What happens when the price changes? "Free for a limited time" and "included with a subscription" can both end.
When should you skip a course?
- You'd rather ship something than finish modules.
- You already write prompts daily and your problem is reliability, which needs evaluation work, not another course.
- The course promises a job title. Titles and skills overlap and change, so read prompt engineer vs AI engineer and the AI engineer interview questions to see what employers actually test.
What are the free alternatives?
This site's own lessons are free. The beginner track covers clarity, context, formatting, roles and iteration, and the intermediate and advanced tracks go further. The prompt library has copy-paste prompts to practise with. That's a biased recommendation since I run the site, so use it alongside whatever course you choose, not instead of checking the others.
